“Mining the large Banking dataset by using K-means, DBSCAN and HAC in WEKA tool”

COMPUTER ENGINEERING

The unprecedented growth of competition in the banking technology has raised the importance of retaining current customers and acquires new customers so that is important analyzing Customer behavior, which is based on bank databases. Data mining refers to the process of retrieving knowledge by discovering novel and relative patterns from large datasets. Analyzing bank databases for analyzing customer behavior is difficult since bank databases are multi-dimensional, comprised of monthly account records and daily transaction records. Clustering the datasets, assessment and the way of expressing customer’s demands and the provinces of requests should be recognized for providing services to the customers, banks, financial and credit institute. Clustering play an important role in data mining. It can make a group of abstract objects into classes of similar objects. In the clustering, firstly partition the set of data into groups based on data similarity and then assigns the labels to the groups. The overall goal of this research work is to evaluate the performance of HAC, K-means and density based clustering (DBSCAN) data mining algorithms by considering the different data sets. HAC is a method of cluster analysis which seeks to build a hierarchy of clusters. It has bottom-up and top-down approach. K-means clustering to partition n observations into K clusters in which each observation belongs to the cluster with the nearest mean. Density based clusters are the dense areas in the data space separated from each other by sparse areas. The above mentioned objective is achieved by WEKA (Waikato Environment for Knowledge Analysis) machine learning tool as an API (application programming interface). This tool for data pre-processing, clustering, classification and visualization. This research present a comparative analysis for various clustering algorithms. In experiments the effectiveness of algorithms is evaluated by comparing the results on the datasets.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.